Community Education and Surveillance on Antibiotics Use Among Young Children in Nepal: A Randomized Controlled Trial

We will design and implement an education intervention targeted to parents of children ages 6 months to 10 years. We will track a number of outcomes related to antibiotics use and health care utilization over 6 months using a surveillance system specifically designed for this purpose.

2023-01-01

2023-12-31

Non-prescribed antibiotics use among children ages 0-10 years

--Health care use (specifically, visits to clinics and hospitals for care rather than a pharmacy)
--Expenditures on antibiotics and other medicines
--Adherence to dosage and duration of use for antibiotics.

The target population consists of parents of children ages 0-10 in Kathmandu, Bhaktapur and Lalitpur districts of the Kathmandu Valley. We randomly select 40 urban wards (from a list of 224 urban wards in these three districts) and randomly assign them to treatment and control arms (20 wards in each). Within the wards, we select households using systematic random sampling. We conduct a baseline survey and intervention in the treatment households. The outcomes will be tracked using a surveillance system designed for the study.

Target population. The target population consists of parents of children ages 0-10 in Kathmandu, Bhaktapur and Lalitpur districts of the Kathmandu Valley. In the federal structure adapted after the new constitution of 2015, these districts are divided into 2 metropolitan cities, 14 urban municipalities, and 3 rural municipalities. The 2 metropolitan cities and the 14 urban municipalities are divided into a total of 224 wards. High penetration of smart phones is a pre-requisite for our intervention (as it involves downloading an App); therefore, we focus on these 224 urban wards where the majority of households own at least one smartphone.

Selection of wards and assignment into groups. To sample this target population, we will use a clustered design with randomization in two stages. In the first stage, we randomly select 40 wards from the 224 wards using a random number generator in the Stata software. In the second stage, we randomly assign the wards into treatment and control (20 wards each), again using a random number generator in Stata. All participants within a ward receive the same intervention (education vs none). This helps avoid contamination across study groups.

We select households for the study using systematic random sampling (and adjust for the size of the ward—thus, the probability of a household being selected for the study—at the time of analysis). Allowing for 10% attrition, which is consistent with the generally low attrition rate in this setting, we will need 960 households (i.e., 480 households in each group; see Sample Size Justification). Therefore, in each ward, we will need 24 households. Using data from the 2021 census, for each ward, we calculate n such that n=total households with at least one child below age 10 divided by 24. In each ward, we pick the first household randomly near the border of the ward and select every nth household until 24 households are recruited in each ward. If a household has more than one child between ages 0-10 years, the youngest child will be the index child (i.e., the child whose antibiotic use, health care use, and medicine costs we will track). The survey team will visit the household, explain the study, seek informed consent, and administer the baseline survey to households that agree to participate. If the selected household does not meet the eligibility criteria or declines to participate, the survey team will visit the adjacent household and follow the same procedure.

We track outcomes using a phone-based app, Antibiotics Surveillance Application (ASA). We will design ASA to support implementation of the education intervention in the treatment group and to track antibiotics, healthcare use, and expenses on antibiotics and other medicine across both treatment and control groups. ASA will serve the following functions:
a) Enroll and collect baseline data (both groups). At enrollment, households will be assigned a unique ID. After enrollment, ASA will be used to conduct the baseline survey.
b) Assist Nurses in providing the “AMR pitch” (treatment group). The nurses in the treatment group will use education module in ASA for the AMR pitch. They will use illustrative digital content (videos, graphs, pictures) to explain the risks of AMR, in addition to face-to-face interaction. We will design, test, and embed videos, flyers, and other illustrative materials to ensure that education to the treatment households is consistently delivered.
c) Enable the households to record and send data on antibiotics and health care use (both groups). The data collectors (who will accompany the nurses to the treatment households and go by themselves in the control households) will demonstrate and train households on the features and user interface of ASA at the conclusion of the visit. Once the educational intervention is completed, households will use the health record module of ASA to record antibiotics use, health care use, and expenses on antibiotics and other medicines after an episode of illness of the index child. From a pre-defined dropdown menu of conditions, symptoms and diagnoses, parents will fill out additional information as applicable. They will also indicate where they sought care, and if they obtained prescription from a physician to buy drugs. Lastly, they will upload pictures of drugs bought and consumed as well as dosage, frequency, duration, and out of pocket costs of such consumption.
d) Send periodic reminders during the 6-month period (treatment group). We will send periodic reminders on rational antibiotic use using videos and text messages to the treatment households using push notification system in ASA. We will send these reminders every two weeks.
e) We provide information on early childhood development to households in both arms. This information is intended to keep households in control group engaged with ASA so that they send information on antibiotics and health care use throughout the study. To keep the two groups comparable in all respects except the education intervention, we provide the information about early childhood development to households in the treatment group as well.

We will randomly select wards using a random number generator in Stata. We will assign the wards into treatment and control arms using Stata as well.

Ward

Yes

40 wards (20 each in treatment and control arms)

960 households

480 households from 20 wards in treatment group. 480 households from 20 wards in the control group.

In our previous study involving standardized patients, community pharmacies provided antibiotics without a physician prescription for pediatric diarrhea and URTI in 60 percent of the cases (average of diarrhea and URTI). In absence of prior data on antibiotics consumption at the household level, we assume that this proportion reflects household-level consumption of non-prescribed antibiotics. A 20 percentage point reduction in this proportion would be meaningful for policy, so we would like to be able to detect this difference at the 5% significance level with at least 80% power. There is no reason to believe that the intervention would have a negative effect—i.e., raise antibiotic consumption—so we base our calculation on a one-tail test. Our intervention will vary at the ward level to avoid contamination, with 20 households selected per ward. We assume a conservative intra-cluster correlation of 0.2. With these assumptions, we need a total of 760 households (380 households in each arm) from 40 wards (20 in each arm) for the final analysis. Our prior experience suggests that non-response and attrition are very low in this setting (in the vicinity of 10%). Nonetheless, we plan on a 20% attrition rate as the study requires participants to use a phone App. Therefore, we will recruit and enroll 480 households in each arm, for a total of 960 households in the study.